Examining the role of testosterone in mediating short-term aggressive responses to social stimuli in a lizard

Abstract

Hormones have been suggested as a key proximate mechanism that organize and maintain consistent individual differences in behavioural traits such as aggression. The steroid hormone testosterone in particular has an important activational role in mediating short-term aggressive responses to social and environmental stimuli within many vertebrate systems. We conducted two complementary experiments designed to investigate the activational relationship between testosterone and aggression in male Egernia whitii, a social lizard species. First, we investigated whether a conspecific aggressive challenge induced a testosterone response and second, we artificially manipulated testosterone concentrations to examine whether this changed aggression levels. We found that at the mean level, plasma T concentration did not appear to be influenced by an aggression challenge. However, there was a slight indication that receiving a challenge may influence intra-individual consistency of plasma T concentrations, with individuals not receiving an aggression challenge maintaining consistency in their circulating testosterone concentrations, while those individuals that received a challenge did not. Manipulating circulating testosterone concentrations had no influence on either mean-level or individual-level aggression. Combined with our previous work, our study adds increasing evidence that the relationship between testosterone and aggression is not straightforward, and promotes the investigation of alternative hormonal pathways and differences in neuro-synthesis and neuroendocrine pathways to account for species variable testosterone - aggression links.

Fig 1. Mean circulating T concentrations (values stated on graph with associated standard error) of male Egernia whitii pre- and post- treatment behavioural challenge for those individuals that did not receive a challenge (light grey bars) and those individuals that did receive a challenge (dark grey bars).

Fig 2. Intra-class correlation coefficients with associated 95% confidence intervals for the consistency of plasma testosterone for those individuals that received a behavioural challenge and those individuals that did not receive a behavioural challenge.

Although the ICC values indicate that those individuals not receiving the challenge remained consistent in their circulating plasma T concentrations, the confidence intervals between the two groups overlap, indicating that they are not different from each other.

Fig 3. Intra-class correlation coefficients with associated 95% confidence intervals for the consistency of aggression pre- to post- treatment across the four treatment groups.

Confidence intervals for all of the groups overlap indicating that the ICC values for each group are not different from each other.